Abstract: Epicenters of non-medical prescription opioid (NMPO) use, heroin use, and drug-related harms (i.e., HIV, HCV, and overdoses [ODs]) are shifting from cities to rural areas; rural young adults (YA) are bearing the brunt of these epidemiologic transitions. The recent HIV outbreak in rural Indiana testifies to the crises these transitions can create. Science, however, has failed to keep pace with these seismic shifts. 15 years after NMPO use began to surge in rural areas, we know little about patterns or determinants of NMPO use, heroin use, HIV, HCV, and ODs among rural YAs. Studies of rural opioid users (regardless of age) have also been limited: with rare exception, this research has focused solely on individual-level predictors. A large body of research guided by the Risk Environment Model (REM) in cities, however, shows that vulnerability to drug use, HIV, and other harms is created by factors operating at multiple levels, including neighborhoods. Almost all methods and measures to study drug use and related harms were created for use in cities, and so developing a new generation of research on rural YA opioid users and their risk environments presents major methodologic and measurement challenges. Targeted sampling, respondent-driven sampling (RDS), and other methods created to recruit urban people who use drugs (PWUD) are most effective when population densities are high, environments are service rich, and public transportation links participants to study sites; these same (urban) features facilitate the in-person survey administration and biosample collection so common in studies of urban PWUD. These recruitment and data collection methods, however, founder in rural areas, which are sparsely populated, have few services for PWUD, and lack public transportation. Applying REM in rural areas also presents measurement challenges, in part because the geographic units at which data are available (e.g., census block groups) are much too large in rural areas to precisely characterize local environments. The proposed mixed-methods R21 study, led by a dual-PI team with rich complementary expertise, will lay the foundations for a new generation of theoretically informed, rigorous, reproducible, high-impact multilevel research on how risk environments influence NMPO and heroin use and HIV, HCV, and OD among YA in multiple US rural counties. It will lay these foundations by leveraging high levels of internet use in the target population and will achieve the following aims: Aim 1. To develop and test the feasibility of using (a) WebRDS, a virtual form of RDS, and (b) a virtual eligibility screener to create a sample (N=200) of YA opioid users in a rural epicenter of NMPO use, heroin use, and related harms (Rowan County, KY). Aim 2. To create and validate novel survey-based and virtual geospatial measures of features of rural risk environments. Aim 3. To develop and pilot (a) an internet survey of the rural risk environment; drug use patterns; HIV and HCV risk; and OD risk and occurrence in the sample; and (b) a system of home-based HIV testing in a subsample. Aim 3.1. To analyze Aim 3 data to inform power calculations for this new generation of high-impact multilevel research.